What do we mean by ‘data’ (1)

• Individual vs aggregated
• Survey vs administrative
• About organisations (ie schools) vs people
• From organisations and young people
• One off studies vs series
• Photography (ie cross-sectional) vs video (longitudinal)
• Our focus today: individual survey microdata

What do we mean by ‘data’ (2)

• Questions asked to people

  • About objective characteristics ie height, weight, age

  • About views and attitude:

    • Attitudes and opinions on crime, gender change
    • Validated instruments (ie mental health)

• External measurement (biosample, genetic, cortisol levels)

• Smart data (from devices ex accelerometer, energy meter)

• Linked administrative records - more on this later

At a glance …

• Cohort studies
• Understanding society
• Linked dataset
• Census

Birth cohort studies

• Follow a sample of individuals from birth until death
• People born during during the same specific period of time in 1958(NCDS), 1970(BCS), 2000 (MCS), 2026 (Generation New Area)
• Wide range of measurements, including education and health
• Within and between cohort analysis
• Managed by the Centre for Longitudinal Studies (UCL)

The Cohort Studies at the CLS

Millennium Cohort Study (MCS)

• ~ 19K children (born between June 2001 and Jan 2003)
• 7 ‘sweeps’: aged 9 months then 3, 5, 7, 11, 14, 17, 23
• Parent and child interviews
• Education, skills, health, truancy, cognitive ability…
• Biological measurements
• … Socio-economic and demographic characteristics
• Data available at the UK Data Service here

Example: Kandola et al (2022)

• Screen-based sedentary behaviour and depression

  • Is a sedentary lifestyle linked to screen and device usage, associated with depression?
  • Does physical activity make a difference?

• 11 341 adolescents from the Millennium Cohort Study

• The main outcome was depressive symptoms:

  • Measured by ‘Moods and Feelings Questionnaire’ (age 14)

• Exposures were frequency of video game, social media, and internet use (age 11).

• Self-reported physical activity (effect modifier)

Results

• IRR: likelihood to be depressed (\(<1\) = less likely, \(>1\) = more likely)

Conclusions

• More frequent gaming at age 11 associated with a lower risk of depressive symptoms at age 14 for boys but not girls.
• More frequent social media use at 11 associated with a higher risk of depressive symptoms in girls but not boys.
• More frequent video game use was consistently associated with fewer depressive synptoms for boys with low physical activity, but not those with high physical activity

Next Steps

• 16,000 people born in 1989-90 followed from age 13-14
• Annual sweeps between ages 14-20, then 25 and 32
• Oversampling of ethnic minorities and deprived schools
• Covers educational attainment and aspirations, health and well-being, and social attitudes
• Linked administrative data available
• Transitions from secondary to post-16 and higher education, family formation, housing, and early labour market entry
• Available at the UK Data Service here

Growing Up in Scotland

• Large-scale, Scotland only cohort study (Website)
• Cohort 1:
  • A sample of children born in 2005-6
  • 11 yearly sweeps. Most recent is Sweep 11 (Age 17-18)
  • 5,217 children at Sweep 1 —> 2,469 at sweep 11
• Cohort 2
  • a sample of children aged around 10 months in 2011.
  • 3 sweeps, most recentis Sweep 3 (Age 5 )
  • 6,127 at Sweep 1 —> 4,434 at Sweep 5
• Growing up in Scotland Series at the UK Data Service

GuS Data and variables

• Mental health (depression, Strengths and Difficulties Questionnaire)

• Diet and sleep; Alcohol, smoking, drug use

• Anti-social behaviour and offending

• Parent physical and mental health

• Linked data

  • Sweep 11 school attainment records - using Scottish Candidate Numbers
  • Sweep 8 Recorded physical activity data

Understanding Society

• Largest longitudinal study of the UK population

• 40K households, 100K individuals initially (2009)

• 17 waves so far: 2009-23. Includes BHPS data 1991-2009

• Ethnic minority boost samples, innovation panel

• Very wide range of topics covered:

  • Employment, income, benefits, savings, debt, and assets
  • Health, well-being, and health behaviour
  • Housing, housing costs, and dwelling characteristics

Young people in Understanding Society

• Youth (ie 10-15yo) self-completion questionnaire
• Usage of smarphone and devices
• Online vs in person socialising
• Usage of apps
• Family support, athmosphere and trust
• Strength and Difficulties questionnaire; self-image
• View about education; schoolwork and bullying

Young people and mental health (1)

Parra-Mujica, Johnson et al (2023)

• Positive association between health for the 16 to 24 and:

  • differences in household income
  • short-term income changes are significant,

• Microsimulation shows that respondents on lower income would benefit the most from increases in their income,

• Possible diminishing effects of income increases on mental health without fiscal and lending reform

• Focusing on the poorest risks neglecting other vulnerable subgroups, ie young workers and single youth

Young people and mental health (2)

Linked administrative datasets

• Records produced when a survey respondent interacts a government organisation

  • hospitalisation, sitting an exam, becoming an employee

• Can be merged with survey data…

  • Where prior consent was obtained from survey respondents

• Entails more stringent conditions of access (Secure access)

• Early days, especially for everyday users but expanding area

Linked data examples

• Understanding Society:
  • Scottish Education Data 2007-2018
  • National Pupil Database (England)
• Millennium Cohort Study
  • Distances to English Grammar Schools, 2012
  • Distances to Current, 1st, 2nd, and 3rd Choice Schools
• Next Steps:
  • Student Loans Company Records 2007-202
  • Individualised Learner Records

Census records

• Fine grained socio-demographic population estimates

• By age group:

  • Deprivation
  • Economic activity
  • Educational status
  • Ethnicity
  • Disability

• Typical application: population estimates for small areas

Residents aged 0-24 as a % of the population by LSOAs , 2021

.

Source

Young People and Gambling Survey

• Attitudes towards and participation in different types of gambling activities

  • Legal and age restricted gambling

• Data controller: Gambling Commission.

• 3,666 pupils aged 11 to 17 yo (school years 7 to 12 (S1 to S6 in Scotland

• You can find it here

Children’s People and Nature Survey

• Population: young people aged 8 -15 yo

• Experience of and views about the natural environment

• Two waves: one in term time, one in holiday time.

• Sample size: about 4,000 per year, 2,000 per wave

• The data is available here

• Topics covered:

  • Wellbeing
  • Time spent outside
  • Quality of outdoor spaces
  • Opportunities and barriers to spending time outside
  • Environmental concern and action
  • Nature connection
  • Countryside code

Smoking, Drinking and Drug Use among Young People

• Information on young people aged 11 to 15 who smoke, drink alcohol or take illegal drugs in England

• It includes information on:

  • prevalence of smoking, drinking and drug taking
  • use of e-cigarettes and emerging drugs
  • types of alcohol and drugs taken
  • how often pupils smoke, drink and take drugs
  • where cigarettes, alcoholic drinks and drugs are obtained
  • pupils’ attitudes to these behaviours

• Sample size: about 17,000 (2023)

Qualitative & deposited research data

• No qualitative equivalent to the major quantitative studies but…
• ESRC grantholders are required to deposit their data
• A growing treasure trove of qualitative (and quantitative ) research data on various topics, including young people
• Often freely accessible via ReShare on the UKDS website, sometimes authorisation of the PI is required

Levels of access

How to access safeguarded data

• Register with the UK Data Service

  • Procedure differs slightly for uk academic users and other users

• Create a project (‘non commercial’, ‘commercial’, ‘teaching’)

• Search, find and order the data

• Add it to your project

• Download it…

• Et voilà.

• A detailed summary is available here

Other levels of access

• Open data: can be downloaded without registration

• ReShare (ie deposited) data: conditions will depend on the data owner; most of the time freely downloadable

• Controlled data:

  • Register
  • Provide a detailed description of your project
  • Apply for Safe Researcher certification
  • Follow the Safe Researcher training
  • Access the data via the SecureLab - ie secure VPN from an institutional computer

Take aways

• Wealth of data
• Survey and non survey
• Education and beyond
• Different ways of searching
• Different tiers of access